language:
- zh
tags:
- roleplay
- multiturn_chat
介绍
基于self-instruct生成的多轮对话roleplay数据，约1k条不同的人格数据和对话
存在问题：
1.基于模型自身生成，所以roleplay存在模型本身价值观融入情况，导致roleplay不够真实，不够准确。
关于我自己：
我是小雨的开发者，小雨是一个情感ai，人格ai，如果对小雨感兴趣的话欢迎支持一下，她目前在bilibili直播，目前我仍在不断的改进 url：https://live.bilibili.com/27357528?broadcast_type=0&is_room_feed=1&spm_id_from=333.999.live_users_card.0.click&live_from=86001
注：
使用本数据集请注明来源
Introduction
This dataset consists of approximately 1,000 instances of multi-turn roleplay conversations generated based on self-instruction. Each instance involves different personalities engaging in dialogue.
Challenges:
The data is generated by the model itself, leading to potential integration of the model's own values into the roleplay scenarios. This may result in roleplays that are not entirely realistic or accurate.
About Myself:
I am the developer of Xiaoyu, an AI specializing in emotion and personality. If you're interested in Xiaoyu, feel free to show your support! She is currently live on Bilibili, and I am continuously working on improvements.
Note:
When using this dataset, please provide proper attribution to the source.
引用
@misc{selfinstruct,
title={Self-Instruct: Aligning Language Model with Self Generated Instructions},
author={Wang, Yizhong and Kordi, Yeganeh and Mishra, Swaroop and Liu, Alisa and Smith, Noah A. and Khashabi, Daniel and Hajishirzi, Hannaneh},
journal={arXiv preprint arXiv:2212.10560},
year={2022}
}